life-cycle assessment of biosolids processing options · peters and lundie, life-cycle assessment...

A P P L I C A T I O N S A N D I M P L E M E N T A T I O N

http://mitpress.mit.edu/JIE Journal of Industrial Ecology 103

� Copyright 2002 by theMassachusetts Institute of Technologyand Yale University

Volume 5, Number 2

Life-Cycle Assessment ofBiosolids Processing OptionsGregory M. Peters and Sven Lundie

Keywords

biosolidsglobal warming potential (GWP)human toxicity potential (HTP)lime amendmentsewage sludgewastewater treatment

Address correspondence to:Gregory PetersSydney Water Corporation115-123 Bathurst St.Sydney NSW 2000Australiagregory.peters@sydneywater.com.auwww.sydneywater.com.au

Summary

Biosolids, also known as sewage sludge, are reusable organic ma-terials separated from sewage during treatment. They can bemanaged in a variety of ways. Different options for biosolids han-dling in Sydney, Australia, are compared in this study using life-cycle assessment. Two key comparisons are made: of system sce-narios (scenario 1 is local dewatering and lime amendment;scenario 2 is a centralized drying system) and of technologies(thermal drying versus lime amendment). The environmental is-sues addressed are energy consumption, global warming potential(GWP), and human toxicity potential (HTP).

Scenario 2 would consume 24% more energy than scenario1. This is due to the additional electricity for pumping and par-ticularly the petrochemical methane that supplements biogas inthe drier. A centralized system using the same technologies asscenario 1 has approximately the same impacts. The GWP andHTP of the different scenarios do not differ significantly.

The assessment of technology choices shows significant differ-ences. The ample supply of endogenous biogas at North Headsewage treatment plant for the drying option allows reductions,relative to the lime-amendment option, of 68% in energy con-sumption, 45% in GWP, and 23% in HTP.

Technology choices have more significant influence on the en-vironmental profile of biosolids processing than does the choiceof system configurations. Controlling variables for environmentalimprovement are the selection of biogas fuel, avoidance of coal-sourced electrical energy, minimization of trucking distances, andraising the solids content of biosolids products.


104 Journal of Industrial Ecology

Introduction

Ecologically sustainable management of thesolid residuals from wastewater treatment,known as biosolids or sewage sludge, is an on-going challenge in Sydney, Australia, as in manyparts of the world. Several components of SydneyWater’s1 legal operating framework require theorganization to act in accordance with the prin-ciples of ecologically sustainable development, asdefined in the Protection of the EnvironmentAdministration Act, 1991. According to theSydney Water Act, 1994, Sydney Water is re-quired to demonstrate a reduction in “the com-bined environmental impact of the per capitaamount of energy and water used . . . and othermaterials . . . discharged.” Environmental life-cycle assessment (LCA) is an obvious means forassessing the potential environmental effects ofmajor capital works, consistent with this stateddesire for holistic assessment. Sydney Water hasdecided to upgrade all biosolids treatment facili-ties to produce “A” grade biosolids (NSWEPA1997), a standard that can be reached with sev-eral technologies. In early 1999, Sydney Waterdecided to apply LCA to the examination of par-ticular proposed upgrades to its biosolids treat-ment infrastructure and potential “A” gradetechnologies. This also presented an opportunityfor Sydney Water to examine the application ofLCA performed to the ISO 14040 series of stan-dards (AS/NZS 1998; ISO 1998, 2000a, b).

The LCA examined two key issues in the fur-ther development of Sydney Water’s biosolidsbusiness: the choice of system configurations, andthermal drying versus lime-amendment technol-ogies.

Several studies have been carried out in thearea of wastewater treatment: Lundin and col-leagues (2000) compared large and small con-ventional wastewater systems with separationsystems, whereas Emmerson and colleagues(1995) analyzed small-scale sewage treatmentprocesses from an LCA perspective. Other LCAstudies on wastewater systems compared alter-native treatment options (Bengtsson et al. 1997;Tillman et al. 1998), different sludge treatmentand recycling processes (Neumayer et al. 1997),and source separation (Dalemo 1996). Applica-tion of LCA to potable water treatment has only

recently begun (Crettaz et al. 1999). Dennisonand colleagues (1998) compared centralized ver-sus decentralized systems and the influence ofcomposting and digesting on the enhancedgreenhouse effect. To our knowledge, this LCAis the first to report on an examination of limeamendment versus thermal drying for biosolidstreatment, and to examine biosolids handling inan Australian context.

LCA Goal and Scope Definition

Goal Definition

The goal of the study is to assess the appli-cation of LCA to Sydney Water planning activ-ities by a holistic environmental analysis of se-lected biosolids processing options for threemajor sewage treatment plants (STPs) discharg-ing treated effluent to the Pacific Ocean (figure1). The approach is prospective or change ori-ented, looking at the upgrades to plants requiredto meet future biosolids production rates. There-fore, the study has natural relevance to the on-going assessment of upgrade options for theocean plants.

Functional Unit

The functional unit of the study is the treat-ment of the mass of biosolids that is expected tobe captured at the three largest plants in Sydney,which serve an estimated population of 2.8 mil-lion people and discharge treated effluent to thesea: North Head, Bondi, and Malabar. The cur-rent estimate of this value for the year 2021, as-suming full primary treatment, is 178 dry tons/day (dt/day).2

System Boundaries

In this study, two functionally equal systemand technology configurations are comparedwith each other:

• System configuration comparison: status quo(scenario 1) versus drying at a central location(scenario 2). Thermal biosolids drying at aproposed central biosolids treatment facil-ity (CBTF) is compared with treatment of


Peters and Lundie, Life-Cycle Assessment of Biosolids Processing Options 105

Figure 1 Map of the Sydney, Australia region showing catchments of the largest STPs.

solids according to current practices. In or-der to examine the consequences of build-ing a central drying facility, calculationswere performed for both a central systemwith drying and pelletization, and for acentral system producing the current ratioof lime-stabilized and dewatered biosolids(these systems are described in more detailin the next section).

• Technology comparison: lime stabilization ver-sus drying of biosolids. Alternative technol-ogies for treatment of sludge were com-pared on a decentralized basis only. Theenvironmental impacts of the applicationof thermal drying and the alternative oflime amendment at North Head STP wereassessed. These systems are described inmore detail in the next section. The func-tional unit was consistent with the com-parison of system configurations: biosolidscapture expected in 2021, currently esti-mated at 70.8 dt/day for North Head.

The foreground3 system boundary of this LCAruns from the outlet of the primary sedimenta-tion tanks to delivery at the land application site.All of the unit operations between these bound-aries are considered, including the ancillary ser-vices of odor abatement and (at Malabar) cogen-eration with biogas. The processes were analyzedat the level of individual unit operations for theforeground system whenever possible. Using na-tional and international databases embedded inthe LCA software, GaBi 3v2 (GaBi 2001), ad-ditional background systems processes were mod-eled.

The inventory for the foreground system in-cluded construction and operational impacts.Process impacts of the background systems wereincluded, but the “cut-off” rule of not includingthe impacts resulting from the installation of thebackground systems was applied.4 Equipment isassumed to have an average operating life spanof 20 years. Because the materials used in con-struction of the plant and equipment (primarily



Figure 2 Conceptual system for scenario 1.

concrete and steel) are recyclable and consideredto reduce environmental impact in other productsystems, the disposal of equipment is not consid-ered in this LCA. Additionally, the material andenergy flows associated with construction areconsiderably smaller than those associated withoperation of the plants, and it is therefore to beexpected that operational issues dominate the to-tal environmental impact of the system relativeto construction and disposal.5

System Configuration ComparisonA conceptual diagram of scenario 1 is given

as figure 2. In this system, the unit operations atthe STPs are substantially the same as those cur-rently in operation: lime amendment without di-gestion at North Head, and digestion and de-watering at Bondi and Malabar. At all sites,vapor scrubbers are used to treat the air drawnfrom the enclosed facilities.

At North Head, the primary settling tank un-derflow is thickened by gravity prior to centrifugedewatering. Physical and microbial stabilizationoccurs via subsequent lime amendment. The bio-solids are trucked to an agricultural applicationsite 250 kilometers (km) away using covered 25

tonne articulated trucks. For the purposes of thisstudy, 50 km of the journey is assumed to be inurban traffic, with the remainder under rural ormotorway conditions.

At Malabar and Bondi, thickening occurs bycentrifugation. Prior to centrifuge dewatering,the biosolids are digested. In the case of Malabar,the biogas is burned in a cogeneration set, andthe waste heat is used to warm the digesters,whereas at Bondi the biogas combustion suppliesthe heat for the digesters via a boiler. The di-gested, dewatered biosolids from these STPs arealso delivered to an agricultural site 250 km away.

It is expected that higher effluent flowratesand suspended solids capture rates will be man-aged at the STPs in this study. The materials forconstruction of the STPs included in this LCAare only those required to upgrade the facilitiesbetween now and 2021. This includes additionalthickening centrifuges and digesters at NorthHead and Malabar, additional dewatering cen-trifuges at Bondi and Malabar, and additionalstorage for treated biosolids at North Head andMalabar. Additional odor scrubbing equipmentis required at Malabar.

Scenario 2 would maintain the use of the



Figure 3 Conceptual system for scenario 2.

thickening and digestion units at Bondi and Mal-abar, with the associated uses of biogas, but theother unit operations are located at a central pro-cessing facility. (Several alternative locations ex-ist at which biosolids handling could be central-ized, if the decision to centralize were made.)Raw sewage would be pumped from North Headto the CBTF for gravity-thickening and diges-tion, both to enhance its stability and to providebiogas fuel for the drier. This fuel is insufficientfor the total heat load of the direct drier and issupplemented with (fossil) natural gas for thisstudy. Figure 3 depicts the system configurationof scenario 2.

Construction materials required for this sys-tem are predominantly concrete and steel. Lo-cating a digester at the CBTF implies the instal-lation of further digestion equipment in additionto that envisaged for scenario 1. The transfer ofsewage to the CBTF would require the laying ofnew concrete-lined, ductile iron pipes betweenthe three STPs and the new plant. A corrosionresistant, fusion-bonded epoxy-coated stainlesssteel pipe would be required for the submergedcrossing of Sydney Harbour by the pipeline fromNorth Head. The CBTF would require new de-watering centrifuges, connected to new directdrying and storage equipment.

The system shown in figure 3 was also ex-amined without the CBTF digester and with thedirect drying unit replaced by a lime-amendmentunit sized to treat the North Head biosolids only.This approach allowed the significance of the actof centralization to be examined (predominantly,the significance of pumping impacts) indepen-dently of the consequences of the differences oftechnology chosen for scenarios 1 and 2.

Technology Comparison: Lime Stabilizationversus DryingFor the comparison of drying technologies,

the process units required for lime stabilizationand drying of biosolids at North Head were com-pared. These are shown in figure 4.

The lime-stabilization system at North Headis exactly as described as part of scenario 1. Thedrying system would be similar to that employedin scenario 2; however, as a consequence of thereduced thermal load of the direct drier, it is pos-sible to supply its entire thermal needs by burn-ing the biogas produced from digestion of thesludge captured at North Head. In fact, surplusbiogas would be produced in the digesters, andthis would be assumed to be used in heating thedigesters, with the remainder flared.



Figure 4 Conceptual alternative for North Head biosolids systems.

The materials required for the lime option arethe same as those described in the comparison ofconfiguration scenarios. For the drying option,thickening centrifuges, digesters, driers, and ad-ditional storage facilities are required.

Comparison of the Systems

The main function of the systems under studyis the treatment of biosolids. Because only oneproduct is considered to leave the plant (treatedbiosolids), all impacts are attributed to it; how-ever, electricity generated from anaerobic diges-tion at Malabar STP is a by-product that crossesthe system boundary. The electricity is exportedto parts of the sewage treatment process otherthan the biosolids processing. This energy exportis taken into account by including the avoidedenvironmental impacts that would otherwise re-sult from coal-fired electricity generation—thesupply alternative to the Malabar STP cogener-ation facility.

Methodological Choices and Assumptions

In this study, several methodological choicesand assumptions were made that might influencethe results. The most relevant choices are listedhere:

• System boundaries. The system starts at theoutlet of the primary sedimentation tanksand continues to the delivery of biosolidsat the land application site. The applica-tion of biosolids is excluded from this studybecause of the complexity of its impacts oninterrelated soil chemical processes.

• Capital equipment. Nonrecurrent (con-struction) impacts associated with long-lived equipment are considered only forthe upgrade of the facilities. Environmen-tal impacts from construction of existingfacilities are excluded.

• Impact assessment models. In this study, im-pact categories and environmental indica-tors have been selected that are most rele-vant to Sydney Water, that is, energyconsumption, global warming potential,and human toxicity potential of air emis-sions.

• Energy consumption. In regards to primaryenergy consumption, all energy flows en-tering the defined system (consumed en-ergy) and all energy flows leaving the de-fined system (produced energy) areconsidered. If energy is produced withinone system and at the same time con-sumed, no energy flows occur across thesystem boundary. Therefore, only the net



energy consumption/production is shown.In the case of Malabar, there is a net pro-duction of electricity under scenario 1, as aconsequence of its generation in the co-generation plant; however, despite theelectricity exported to other parts of theSTP, the biosolids process is a net con-sumer of energy due to the high energy de-mand of the trucking operations. By con-trast, in scenario 2, this trucking occursfrom the CBTF site, and the additional en-ergy saved at Malabar because of the ab-sence of dewatering operations makes thebiosolids process a net producer of bothelectricity and energy.

• Greenhouse effect. Carbon dioxide and ni-trous oxide emissions resulting from thecombustion of fossil fuels contribute toglobal warming and other climate changeeffects; however, carbon dioxide emissionsfrom the microbial degradation of biosol-ids, or the combustion of biogas, merelycomplete a biological cycle that beginswith the conversion of atmospheric carbondioxide to biomass by photosynthesis incrops. These latter carbon emissions arenot derived from the lithosphere but returncarbon to the atmosphere. Therefore, car-bon dioxide emissions resulting from thecombustion or emission of gases generatedby the metabolizing of biosolids are ex-cluded from the calculation of contribu-tions to the enhanced greenhouse effect inthis LCA, consistent with the approachused by other LCA practitioners (USEPA1997).

Life-Cycle Inventory DataCollection

After validation of annual mass and energybalances, the data were related to unit processesand the functional unit (see the previous descrip-tion of the functional unit and system bound-aries). Finally, the data were aggregated for thelife-cycle impact assessment phase (described inthe next section). This procedure was performedseveral times. The iterative character of the pro-cess allowed us to determine which issues weremost significant prior to finalization of the cal-

culations, which permitted efficient allocation ofefforts to improve data quality when necessary.

The data on these units is contained in sev-eral statutory (Sydney Water 1999b) and plan-ning documents (e.g., Sydney Water 1996,1998a, 1998b, 1999a), supplemented when nec-essary with supplier data (e.g., Peters 1999) andliterature values (Kogan and Torres 1996). Thedata is considered prospective in nature—basedon estimates of future necessary plant and equip-ment using contemporary operational observa-tions. Australian inventory data for the inclusionof background systems was obtained from Grantand colleagues (1999a).

Planning data is supplied to Sydney Water byits consultants on the basis that the cost esti-mates are accurate to within 25% of a statedvalue. Given that inaccuracy can occur both inthe sizing and performance of equipment and inthe pricing of the equipment chosen, the sizingand performance data should be accurate tomuch better than 25%. Performance data fromsuppliers can be expected to accurately reflectoptimal installation of their equipment. Life-cycle inventory and literature data are in thepublic domain yet may possibly be weaker be-cause of their lack of specificity. On the otherhand, the fact that such accuracy can affect eachoption being compared in LCA in a similar man-ner reduces its influence on the research out-comes, resulting in a “standard” or regular error(rather than a random error). A complete dis-cussion of errors and error propagation could gomuch further than this (e.g., Finnveden et al.2000) but is beyond the scope of this study.

Life-Cycle Impact Assessment

Selection of Impact Categories andEnvironmental Indicators

The environmental indicator and impactcategories chosen for this study are energy con-sumption, global warming potential, and humantoxicity potential (HTP). These were chosen onthe basis that they are most relevant to the par-ticular systems undergoing comparison. Impor-tantly, they are also three categories that havebeen the subject of considerable scientific debateand investigation.



Although not strictly an environmental im-pact category, primary energy consumption isuseful as an indicator of the process intensity andthe use of nonrenewable resources. It can alsoprovide useful explanatory data when examiningcontributions to the enhanced greenhouse effectand is in any case a prerequisite for the evalua-tion of the global warming potential of processsystems.

Global warming is obviously of internationaland local interest, given Sydney Water’s status asa major consumer of electrical energy and Aus-tralia’s poor per capita performance in this area.Global warming potential is usually evaluated ona 20, 100, or 500 year timescale. For this study,the middle of those three timescales has beenselected.

HTP of airborne contaminants is of consid-erable interest in urban environments such asSydney Water’s area of operations. HTP has beenstudied in depth by Heijungs and colleagues(1992), Cowan and colleagues (1995), Lynch(1995), Guinee and colleagues (1996a, 1996b),Gorree and colleagues (1999), Jolliet (1996),Hauschild and Wenzel (1998a, 1998b), RIVMand colleagues (1998), and Huijbregts (1999).

The Australian HTP model takes into ac-count 182 potentially toxic substances and theireffects on humans (Huijbregts et al. 2001). Thepotential effect depends on the actual emissionand the fate of the specific substances emitted tothe environment. The fate of chemical sub-stances depends on specific characteristics suchas degradation rate, bioaccumulation, evapora-tion, and deposition. Characterization factorsrelative to 1,4-dichlorobenzene have been cal-culated on the basis of the Uniform System forthe Evaluation of Substances in the context ofLCA (USES LCA) model developed by Hu-ijbregts (1999). The model has been adapted toAustralian conditions.6 HTP factors for 20 sub-stances relevant to this study are shown in table1. These factors consider initial emissions to airand their direct potential impact on humanhealth only. An infinite time horizon has beenchosen.

Data Quality and Sensitivity Analysis

Two types of sensitivity analysis were appliedto this LCA. The first type consists of testing the

assumptions about the system by variation of theconfiguration and/or boundaries. This type ofanalysis has the potential to produce large vari-ations in results. In this study, this type of sen-sitivity analysis is performed by including or ex-cluding biogas from the energy accounts and byexamining scenario 2 with both the alternative(drying) system and the mix of stabilization tech-nologies used in scenario 1. The results of theseassessments are discussed alongside the other re-sults in the next section.

The second type of sensitivity analysis in-volves varying input values for a particular sys-tem. Ideally, each data element in a life-cycle in-ventory should be collected with an error marginreflecting the level of certainty associated witheach datum. This is not appropriate to a scanninglevel LCA nor a prospective LCA such as this.In the absence of retrospective, measured data,it is more appropriate to apply a standard errorto the output. As discussed previously in the con-text of data collection, the planning documentson which most of the inventory for this LCA relyare accurate to within 25% of the estimated dol-lar cost value, and it is therefore considered ap-propriate to apply a standard 25% error toleranceto the output of the assessments.

Results and Discussion

System Configurations

The scenarios are compared on the bases ofenergy consumption, global warming potential,and HTP.

Energy ConsumptionThe net energy consumption of each of the

options is of the same order of magnitude rangingfrom 17.9 to 23.0 terajoules per year (TJ/yr).7

Scenario 1 draws 18.5 TJ/yr compared with 23.0TJ/yr for scenario 2 with drying and 17.9 TJ/yrwithout drying. Note that in scenario 1, there isno CBTF, and in scenario 2, the energy requiredfor pumping the solids from North Head to theCBTF is counted in the CTBF figure. Conse-quently, there is no bar in figure 7 for the CBTFin scenario 1, nor for North Head in scenario 2.This pattern is repeated in figures 5 and 6.

It should be noted that many of the burdensassociated with biosolids treatment at North



Table 1 Human toxicity equivalence factors of selected chemicals

SubstanceEquivalence factors

(1,4-dichlorobenzene equivalent) SubstanceEquivalence factors

(1,4-dichlorobenzene equivalent)

Ammonia 1.56E�02 Mercury 1.23E�03Arsenic 2.95E�04 Nickel 3.03E�03Benzene 1.62E�02 NO2 5.47E�02Cadmium 1.24E�04 Phenol 2.28E�02Chromium (III) 2.97E�01 Selenium 8.12E�03Copper 3.66E�02 SO2 7.95E�03Ethylbenzene 4.66E�02 Styrene 1.67E�03H2S 1.80E�02 Toluene 1.68E�02Hydrogen chloride 7.31E�02 Vanadium 9.42E�02Lead 2.82E�01 Zinc 9.08E�00

Figure 5 Global warming potential for different scenarios.

Head, Malabar, and Bondi are merely transferredto the CBTF in scenario 2. This is why the gen-eration of electricity appears only as an energycredit at Malabar in scenario 2: The main con-sumers of electrical power in the current Malabarbiosolids processing (dewatering centrifuges) aretransferred to the CBTF with the dewateringunit operation.

From an overall energy accounting perspec-tive, the biogas produced and used within theplant does not appear in figure 7 because it doesnot cross the system boundary. The volume ofbiogas produced from North Head biosolids isnot sufficient for the drying of the biosolids fromall three plants. Consequently, the additional en-

ergy drawn by scenario 2 with its drying systemis predominantly the result of the natural gas(methane) required.

The differences between the energy consump-tion of the alternatives vary in significance. Withdrying, scenario 2 uses 24% more energy thanscenario 1 (just less than the 25% error toler-ance). Scenario 2 without drying is not signifi-cantly more energy efficient than scenario 1 (us-ing only 3% less energy). This is the consequenceof the avoidance of dewatering centrifuges atBondi in this option. Dewatering of the solidsoccurs instead at the CBTF, where there is suf-ficient capacity to dewater these solids withoutthe installation of an additional centrifuge. The



Figure 6 Human toxicity potential under different scenarios.

Figure 7 Energy consumption under different scenarios.

energy savings that result from centralizing thedewatering function are of the same scale as theadditional energy required to pump the biosolidsto the CBTF. Scenario 2 with drying draws 28%more energy than scenario 2 without drying, sowe can see that the drying technology, rather

than the choice of a central facility, is more in-fluential.

The most significant contributor to the en-ergy consumption of scenario 1 is the fuel for thediesel trucks transporting the biosolids and lime.Trucking represents 36% of the total energy con-



sumed by the biosolids handling processes atthese three major ocean plants. This suggeststhat reducing trucking distances between STPsand biosolids users, and reducing biosolids mois-ture content, are relatively effective ways to im-prove the energy efficiency and environmentalperformance of biosolids handling.

Greenhouse EffectBiogas is a by-product of biosolids stabiliza-

tion when digestion is used. The biogas is eitherflared or used productively to generate electricity,to dry and pelletize the solids, or to heat the di-gesters whence it came. In any case, it is con-verted to carbon dioxide, which has a lowerglobal warming potential than uncombusted bi-ogas.

If carbon dioxide emissions from biogas com-bustion are excluded in accordance with themethodological choices discussed earlier, the re-sults of the assessment are as follows: Scenario 1makes a contribution to the enhanced green-house effect of 17.6 � 106 kilograms (kg) carbondioxide equivalents, compared with 18.3 � 106

kg carbon dioxide equivalents for scenario 2 withdrying. This is an insignificant increase of 4%over scenario 1. Scenario 2 without drying resultsin the emission of 17.1 � 106 kg carbon dioxideequivalents (figure 5).

The results with regard to global warming al-ter if carbon dioxide emissions that leave the sys-tem boundary are counted in the life-cycle in-ventory. If carbon dioxide from drying at theCBTF is included, the choice of scenario 2 causesan 82% increase in the contribution to the en-hanced greenhouse effect to 78.9 � 106 kg car-bon dioxide equivalents from the scenario 1value of 40.6 � 106 kg carbon dioxide equiva-lents.

Human Toxicity PotentialIn both scenarios, most of the toxic sub-

stances emitted to the environment are airbornecontaminants released as a result of the high en-ergy intensity of the processes. Therefore, theemphasis of the HTP assessment is on all sub-stances released to air (as listed in table 1).

The combined potential for human toxicityvia airborne contaminants is lowest for scenario2 without drying (4.14 � 104 kg 1,4-dichloro-

benzene equivalent/yr) and insignificantly higherfor the other two options (4.31 � 104 and 4.50� 104 kg 1,4-dichlorobenzene equivalent/yr forscenario 1 and scenario 2 with drying, respec-tively). The results are influenced by several fac-tors:

• At Malabar, credits occur as a result ofelectricity generation. Australian electric-ity production usually contributes signifi-cantly to heavy metal emissions via thesmokestacks of coal-fired power generators.At Malabar STP, electricity is generatedand exported from the boundary of the bio-solids handling system to other parts of theSTP. This represents avoided electricalgeneration in other parts of New SouthWales, Australia.

• Trucking is the major contributor to HTPof the North Head (lime) and CBTF (with-out drying) systems. In these cases, approx-imately two-thirds of the toxic emissionsare caused by transportation.

• In scenario 2 with the drying process at theCBTF, the major contributions to toxicemissions come from drying process (33%),dewatering (29%), and trucking (about16%).

• The result is predicated on the absence ofheavy metals in biogas and natural gas, andthe heavy-metal-contaminated fuel used inthe majority of New South Wales’ electri-cal generation. Approximately 5% of NewSouth Wales’s electricity is of hydroelectricorigin, with the remainder predominantlycoal-sourced (Bush et al. 1999).

Lime Amendment versus Thermal Dryingand Pelletization Technologies

The technology comparison of lime amend-ment versus thermal drying is made on the basisof the same key indicators used in the compari-son of system scenarios.

Energy ConsumptionThe following discussion considers the total

annual energy demand for lime amendment andthermal drying. The values calculated for energyconsumption of the lime and drying systems for



Figure 8 Energy consumption: drying versus lime at North Head.

North Head biosolids treatment are 133 and 42.1TJ/yr, as shown in figure 8. (The results areshown in megajoules per year.)

The comments made in connection with theimpact of trucking on the overall energy con-sumption of the different system scenarios aresupported dramatically in this figure. The energyconsumed by trucking greater masses of biosolidsgenerated by lime amendment, and the lime re-quired for the process, result in the total energyconsumption of the drying option being only32% of the amount required for lime amend-ment. Because transportation of limestone to theplant represents only 8% of the trucked materialby mass, the most significant issues are the dis-tance the biosolids are trucked and their mass.The pelletized and the lime-amended biosolidsare transported to the same sites in this compar-ative study, so the key difference is the moisturecontent of the biosolid product. Pelletized bio-solids can be assumed to reach 92% solid materialby mass, whereas lime-amended biosolids are typ-ically only 34% solids.

Note that consistent with figure 7, figure 8does not include endogenous biogas used bene-

ficially as energy consumed, because it is gen-erated within the system boundary. If it wereincluded, it would increase the energy consump-tion of the North Head drying option from 42.1TJ/yr to 283 TJ/yr in this case.

Greenhouse EffectAll greenhouse gas emissions of the two types

of technology are included here. In keeping withour approach to the assessment of the impact ofconfigurations on the overall global warming po-tential of the biosolids handling, figure 9 showsthe results of the assessment of greenhouse gasemissions at North Head, excluding biogas com-bustion from the calculations. The contributionto the enhanced greenhouse effect of the dryingoption is 45% better than the lime-amendmentoption: 8.75 � 106 and 1.59 � 107 kg carbondioxideequivalents, respectively. This is a signifi-cant difference, although relatively small com-pared to the difference in energy consumption.This is a consequence of the difference in green-house intensity of fuel consumption by dieselmotors and coal-fired power generators. It is sig-nificant that the fuel sources of the drier are the



Figure 9 Global warming potential: drying versus lime at North Head.

digesters within the system boundary. If naturalgas were used to fire the drier, an additional 9.28� 106 kg carbon dioxide equivalents would beemitted, making the lime-amendment optionclearly preferable in terms of its contribution tothe enhanced greenhouse effect.

Human Toxicity PotentialHTP includes relevant substances released to

the atmosphere (as listed in table 1).Consistent with the overall analysis of HTP

in the examination of different configurations,figure 10 shows that the HTP of the lime-amendment process is 21% higher than for thethermal drying process. The most significantcontribution to human toxicity is made by truck-ing operations, primarily as a consequence of thelarge quantities of benzene emitted to the at-mosphere.

The toxicity associated with drying is a con-sequence of the electrical consumption of thepumps involved in maintaining the mixed stateof the digester and of decanting it. Drying alsoinvolves electrical power (estimated at 240 kW)to rotate the drier drum and provide other an-cillary services.

Additional Comparisons Drawn

Resource Consumption: North Head(Drying)Data exist on the quantity of nonrenewable

fuels used in the processes of generating electric-ity, building, and driving trucks. This is presentedraw (on a nonnormalized basis) in figure 10 alongwith the materials used during construction. Thisindicates that the operation of the North Headdrying option uses far more resources than theconstruction of the plant to do it. The figure isbased on one year of operation. Averaging thequantity of materials over the 20 years of theiroperational life reduces the total annual con-sumption of construction materials to 5% of thematerials used in the operation of the plant. Di-gestion is the most material-intensive part of theplant as a consequence of the scale of the units,representing approximately 90% of the total con-struction material budget (figure 11).8

In interpreting this data, it is worth bearingin mind that this study examines the incrementaladditions to the infrastructure required to oper-ate the systems at full primary level. The majorityof the biosolids plant required for this has alreadybeen constructed in order to treat the solids re-



Figure 10 Human toxicity potential: drying versus lime at North Head.

Figure 11 Resource depletion at North Head (drying): operation versus construction.



Figure 12 Energy consumption by transportation in different scenarios.

moved under the current system of less than pri-mary treatment.

Energy Consumed by TransportationThe energy required for pumping thickened

biosolids at Malabar and Bondi in scenario 2 isaccounted for in the total figures for these plants.Figure 12 indicates the relative significance ofthese amounts in the overall energy budgets. Us-ing thermal drying processes decreases the trans-port energy budget significantly whether it isused to treat the biosolids from the three majorocean plants, or merely North Head. The energyinvolved in transportation of dewatered or limedbiosolids is a major component of the energybudget of the associated biosolids treatment pro-cesses.

Conclusions

General Conclusions

The environmental impacts explored by thisLCA represent only some of the factors that Syd-ney Water must consider as it plans to upgradebiosolids facilities. In terms of regional and global

issues, this LCA indicates that whether SydneyWater makes a choice to centralize or not willnot effect the environmental outcomes as muchas the technology choices. It also suggests thatthe choice of treatment technologies other thanbiogas-fueled thermal drying would result in ad-ditional environmental impact.

The life-cycle resource consumption of thedrying option is dominated by the operation ofthe system, rather than its construction. The re-sources are predominantly consumed by energyproduction for the plant. As previously stated,the choice of biosolids processing technology ismore significant than the choice of system con-figuration scenarios.

The largest proportion of the energy requiredby transportation activities is consumed in truck-ing. The drying options (whether installed cen-trally or merely at North Head STP) require theleast energy for transportation.

A key issue in improving the environmentalprofile of biosolids handling is the avoidance ofcoal-sourced electrical energy. Selection of re-newable energy sources such as biogas, andcleaner energy sources such as natural gas, willimprove the environmental performance of bio-



solids operations by reducing both the carbon in-tensity of energy use, and the toxicity of emissionby-products. This has been shown for both thedecision between drying and lime amendmentand the choice of whether or not to installbiogas-fueled cogeneration equipment at NorthHead or the CBTF.

Alternative System Scenarios

Scenario 2 would not use significantly moreenergy than scenario 1. Additional energy is re-quired for pumping the settled biosolids and forthe operation of the drier, but this is partiallyoffset by the reduction in energy for transporta-tion of treated biosolids and quicklime. This out-come is not controlled by the fact of centraliza-tion, but has to do with the choice of drying andpelletizing technology for all the biosolids in-stead of lime amendment at North Head and de-watering at Bondi and Malabar.

The combustion of biogas was not includedin calculations of global warming potential forthe reasons outlined in the initial discussion ofmethodological choices. Taking this into ac-count, the contribution to the enhanced green-house effect of the different options does not varysignificantly.

The HTP of scenario 2 without drying is mar-ginally lower than that of the other systems, thatis, 4% less than scenario 1 and 8% less than sce-nario 2 with drying. This is a consequence of thechoice of drying technology, which allows a pro-portion of the trucking activities to be avoided.

Lime Amendment versus Thermal Dryingand Pelletization Technologies

The energy consumption of the North Headbiosolids processing system with thermal dryingis only 68% less than that of the lime-amendment option.

The drying option is approximately 45% bet-ter than the lime-amendment option in terms ofits contribution to the enhanced greenhouse ef-fect.

Selection of the drying option would result ina 21% improvement in HTP compared with thelime-amendment option at North Head STP.

LCA as a Decision Making Support Tool

Although it is not the role of this report todetermine developments in Sydney Water’s de-cision making processes, this prospective LCAstudy allows us to assess the potential value ofLCA in supporting planning activities in the fu-ture.

LCA can provide unique contributions tostrategic planning for Sydney Water because itconsiders diverse environmental impacts relatedto wastewater treatment systems and processes.Environmental impacts are taken into accountwhether they occur on- or off-site. Hence, theholistic approach allows a comprehensive envi-ronmental assessment that prevents against“problem shifting,” that is, the transfer of envi-ronmental problems from one part of the tech-nical system or the environment to another.

The available process data, impact models,and software used are adequate to this kind ofstudy. Specialized LCA software (e.g., GaBi3) isvaluable, particularly at the inventory and im-pact stage. It allows easy access to large quantitiesof data about important background systems thatwould otherwise have to be collected for eachstudy.

The results of this study supply new insightsinto the environmental impact of the processesstudied. The recognition of the significance oftrucking in energy consumption, global warming,and HTP issues, for example, is ensured by theLCA methodology. Further studies might con-sider more environmental indicators and impactcategories. Each LCA is specific to the cases un-der analysis, so it may be expected that futureLCAs in other parts of the business will offerother new insights. LCA can be used to improve“life-cycle thinking” in Sydney Water’s planningactivities, taking into account the entire life spanof a plant, and the other processes in its supplychain. This is essential in order to properly assessthe ecological sustainability of developments.

Notes

1. Sydney Water is a statutory state-owned corpo-ration responsible for water supply, wastewatertreatment and some stormwater services for thecommunities of Sydney, the Blue Mountains, andIllawara, Australia.



2. In all instances in this article, “ton” refers to“metric ton”: 1 metric ton � 1 Mg (SI unit) �

1.1 short ton � 2,205 lbs.3. “Foreground systems” refers to the alternative sys-

tems under examination in this study, as shownin figures 2–4. The “background” system “com-prises all other processes which interact directlywith the foreground system, usually supplying ma-terial or energy to the foreground or receivingmaterial or energy from it” (Clift et al. 1999) andare not the domain of the alternatives being ex-amined.

4. This is generally justified on the basis that (1) formost supply chains, the proportion of the outputof the chain that would be directed to the fore-ground system is minimal; and (2) the impact ofplant construction activities is generally muchless than the impacts of their operations.

5. This has been qualitatively confirmed by Otomaand colleagues (1997).

6. The Australian human toxicity model is based onthe USES LCA model (Huijbregts 1999). Thismodel has been adapted to Australian conditionsby integrating Australian environmental param-eters, human characteristics (e.g., daily intake offood), input parameters for fate analysis, and hu-man exposure assessment and weighting factorsfor aggregation of the risk characterization ratio(Huijbregts et al. 2001).

7. 1 terajoule � 1012 J � 9.48 � 106 BTU.8. The construction, operation, and demolition of

small-scale sewage-treatment processes have beenanalyzed by Emmerson and colleagues (1995) andLundin and colleagues (2000). The energy con-sumption of construction is a significant impacton the overall environmental profile in both stud-ies. Emmerson and colleagues (1995) determinedthat the energy consumption ranges from 5% to50% for the construction of the plants, whereasthe contribution to climate change is significantlylower (less than 15%). Resource consumption,however, has not been considered in any of thesestudies.

Acknowledgments

The authors thank Rachel Clarke (SydneyWater) and Christina Dimova (CWWT) for sig-nificant contributions to the data collection andmodeling involved in this project. We also thankDavid Gough (Sydney Water) and Ian Hammer-ton (Sydney Water) for their cooperation andreview of the report.

References

AS/NZS. 1998. AS/NSZ ISO 14040:1998 Environmen-tal management—life cycle assessment—principlesand framework. Standards Australia and Stan-dards New Zealand. Published jointly by Home-bush, Australia; Wellington, New Zealand.

Bengtsson, M., M. Lundin, and S. Molander. 1997. Lifecycle assessment of wastewater systems. Gothen-burg, Sweden: Technical Environmental Plan-ning, Chalmers Tekniska Hogskolan.

Bush, S., A. Dickson, J. Harman, and J. Anderson.1999. Australian energy: Market developments andprojections to 2014–2015. ABARE Research Re-port 99.4. Canberra, Australia: Australian Bureauof Agricultural and Resource Economics.

Clift, R., R. Frischknecht, G. Huppes, A.-M. Tillman,and B. Weidema. 1999. SETAC working groups1993–1998. SETAC—Europe News 10(3): 17.

Cowan, C. E., D. Mackay, T. C. J. Feijtel, D. van deMeent, A. Di Guardo, J. Davies, and N. Mackay.1995. The multi-media fate model: A vital tool forpredicting the fate of chemicals. Pensacola, FL: So-ciety for Environmental Toxicology and Chem-istry–USA.

Crettaz, P., O. Jolliet, J.-M. Cuanillon, and S. Orlando.1999. Life cycle assessment of drinking water andrain water for toilets flushing. Aqua 48(3): 73.

Dalemo, M. 1996. The modeling of an anaerobic digestionplant and a sewage plant in the ORWARE simulationmodel. Report 213. Uppsala, Sweden: SwedishUniversity of Agricultural Science, Departmentof Agricultural Engineering.

Dennison, F. J., A. Azapagic, R. Clift, and J. S. Col-bourne. 1998. Assessing management options forwastewater treatment works in the context of lifecycle assessment. Water Science and Technology38(11): 23–30.

Emmerson, R. H. C., G. K. Morse, J. N. Lester, andD. R. Edge. 1995. Life-cycle analysis of small scalesewage-treatment processes. Journal of the Char-tered Institution of Water and Environment Man-agement 9: 317–325.

Finnveden, G., J. Johansson, P. Lind, and A. Moberg.2000. Life cycle assessments of energy from solidwaste. Ursvik, Sweden: Stockholm Universitet/Systemekologi och FOA.

GaBi. 2001. http://www.gabi-software.com/englisch/home_englisch.shtml. Accessed February 2001.

Gorree, M., J. B. Guinee, R. Heijungs, G. Huppes, R.Kleijn, H. A. Udo de Haes, E. van der Voet, andM. N. Wrisberg. 1999. Environmental life cycle as-sessment: Backgrounds. Leiden, the Netherlands:Centre for Environmental Science, Leiden Uni-versity.



Grant, T., C. Dimova, A. Tabor, and J. Todd. 1999a.Life cycle assessment Australian data inventory pro-ject: Summary report. Release 1.1. Melbourne,Australia: Royal Melbourne Institute of Tech-nology.

Guinee, J. B., R. Heijungs, L. van Oers, D. van deMeent, T. Vermeire, and M. Rikken. 1996a. LCAimpact assessment of toxic releases. Generic model-ling of fate, exposure and effect for ecosystems andhuman beings with data for about 100 chemicals. Re-port No. 1996/21. The Hague, the Netherlands:Ministry of Housing, Spatial Planning, and theEnvironment (VROM).

Guinee, J. B., R. Heijungs, L. van Oers, A. WegenerSleeswijk, D. van de Meent, T. Vermeire, and M.Rikken. 1996b. USES: Uniform system for theevaluation of substances. Inclusion of fate inLCA characterisation of toxic releases. ApplyingUSES 1.0. International Journal of Life Cycle As-sessment 1(3): 133–138.

Hauschild, M. Z. and H. Wenzel. 1998a. Ecotoxicitytoxicity as a criterion in the environmental as-sessment of products. In Scientific backgrounds toenvironmental assessment of products 2. London:Chapman and Hall.

Hauschild, M. Z. and H. Wenzel. 1998b. Human tox-icity as a criterion in the environmental assess-ment of products. In Scientific backgrounds to en-vironmental assessment of products 2. London:Chapman and Hall.

Heijungs, R., J. B. Guinee, G. Huppes, R. Lankreijer,H. A. Udo de Haes, and A. Wegener Sleeswijk.1992. Environmental life cycle assessment of prod-ucts: Guide and background. Leiden, the Nether-lands: Centre for Environmental Science, LeidenUniversity.

Huijbregts, M. 1999. Priority assessment of toxic substancesin the frame of LCA-development and application ofthe multi-media fate, exposure and effect model USES-LCA. Amsterdam: Interfaculty Department of En-vironmental Science, Faculty of EnvironmentalSciences, University of Amsterdam.

Huijbregts, M., S. Lundie, and C. Dimova. 2001. Aus-tralian life cycle impact assessment of toxic sub-stances. In CRCWMPC 2001: Australian life cycleimpact assessment project (project 046003). Sydney:Cooperative Research Centre for Waste Manage-ment and Pollution Control.

ISO (International Organisation for Standardization).1998. ISO 14041 Environmental management—Life cycle assessment—Goal and scope definitionand life cycle inventory analysis. Geneva: ISO.

ISO. 2000a. ISO 14042 Environmental management—Life cycle assessment—Life cycle impact assessment.Geneva: ISO.

ISO. 2000b. ISO 14043 Environmental management—Life cycle assessment—Life cycle interpretation. Ge-neva: ISO.

Jolliet, O. 1996. Impact assessment of human and eco-toxicity in life cycle assessment. In Towards amethodology for life cycle impact assessment: Part 4,edited by H. A. Udo de Haes. Brussels: Societyfor Environmental Toxicology and Chemistry–Europe.

Kogan, V. and E. M. Torres. 1996. Criteria pollutantsand air toxic contaminants emissions from com-bustion sources at wastewater treatment plants.In Proceedings of WEFTEC ’96, the 69th annualconference and exposition of the Water EnvironmentFederation. Virginia: Water Environment Feder-ation, 375–383.

Lundin, M., M. Bengtson, and S. Molander. 2000. Lifecycle assessment of wastewater systems: Influenceof system boundaries and scale on calculated en-vironmental loads. Environmental Science Tech-nology 34: 180–186.

Lynch, M. R. 1995. Procedures for assessing the environ-mental fate and eco-toxicity of pesticides. Brussels:Society for Environmental Toxicology andChemistry–Europe.

Neumayer, R., R. Dietrich, and H. Steinmuller. 1997.Life cycle assessment of sewage sludge treatment.In Proceedings of the fifth SETAC annual confer-ence. Brussels: Society of Environmental Toxi-cology and Chemistry.

NSW. 1991. Protection of the Environment Adminis-tration Act. No. 60. New South Wales. www.epa.gov.au/legal/envacts.htm. Accessed February2001.

NSW. 1994. Sydney Water Act. No. 88. New SouthWales. www.epa.gov.au/legal/envacts.htm. Ac-cessed February 2001.

NSW. 1994. Sydney Water Act. No. 88. Part 6, Di-vision 10, Section 22(3)e, p. 21. New SouthWales. www.epa.gov.au/legal/envacts.htm. Ac-cessed February 2001.

NSWEPA (New South Wales Environment Protec-tion Authority). 1997. Environmental guidelines:Use and disposal of biosolids products. Chatswood,Australia: New South Wales Environment Pro-tection Authority.

Otoma, S., Y. Mori, A. Terazono, T. Aso, and R. Sa-meshima. 1997. Estimation of energy recoveryand reduction of CO2 emissions in municipalsolid waste power generation. Resources, Conser-vation and Recycling 20: 95–117.

Peters, G. M. 1999. Personal communication withG. M. Peters, Sydney Water Corporation, regard-ing Andritz sludge drying systems product infor-mation (Andritz reference 0125) supplied by B.



Welungoda, Andritz Pty Ltd. Dandenong, Aus-tralia. 26 July.

RIVM (National Institute of Public Health and theEnvironment), VROM (Ministry of Housing,Spatial, Planning and the Environment), andVWS (Ministry of Health, Welfare and Sport).1998. Uniform system for the evaluation of sub-stances 2.0 (USES 2.0). RIVM Report679102044. The Netherlands: RIVM, VROM,and VWS.

Sydney Water. 1996. Report on cogeneration facility atMalabar sewage treatment plant. Contract 12319,prepared by Sinclair Knight Merz, July. Sydney:Sydney Water Corporation.

Sydney Water. 1998a. Malabar STP PRP—Options re-port, 2 volumes. Contract 15087, prepared bySinclair Knight Merz and CDM Inc. November.Sydney: Sydney Water Corporation.

Sydney Water. 1998b. Bondi STP PRP—Options report.2 volumes. Contract 15087, prepared by SinclairKnight Merz and CDM Inc. November. Sydney:Sydney Water Corporation.

Sydney Water. 1999a. Strategic analysis of consolidated

options, 2 volumes. Contract 15087, prepared bySinclair Knight Merz and CDM Inc. June. Syd-ney: Sydney Water Corporation

Sydney Water. 1999b. Annual environment report 1998.Tillman, A.-M., M. Svinby, and H. Lundstrom. 1998.

Life cycle assessment of municipal waste watersystems. International Journal of Life Cycle Assess-ment 3(3): 145–157.

USEPA (U.S. Environmental Protection Agency).1997. Greenhouse gas emissions from municipalwaste management. EPA contract 68-W6-0029.Washington, DC: U.S. Environmental Protec-tion Agency, Office of Solid Waste and Office ofPolicy, Planning, and Evaluation.

About the Authors

Gregory M. Peters is an Environmental Engineerwith the Sydney Water Corporation.

Sven Lundie Ph.D. is the LCA Discipline Leaderat the Centre for Water and Waste Technology at theUniversity of New South Wales.

life-cycle assessment of biosolids processing options · peters and lundie, life-cycle assessment...

Documents